Short-term Load Prediction of Integrated Energy System Based on MCNN-LSTM Neural Network

The high coupling characteristics between various energy of integrated energy system (IES) and the uncertainty of each link of power-grid-load make the dynamic process of IES more complex than that of the traditional single-load system, which brings great difficulties to the effective operation of IES. To solve the above problems, this paper proposes a method for short term multivariate load prediction based on deep neural network. Firstly, the correlation between each single load is used to divide the categories, and the load data at the same time is reorganized into an image mode to maximize the use of correlation; then, the multi-channel convolutional neural networks (MCNN) are used to extract and fuse the features of the images, and the fused feature data is input into the long short-term memory (LSTM) neural network to realize the prediction of multi-load data from the time level; finally, the feasibility of the proposed method is verified by an actual example analysis.